modules light-soaking tony sample...preconditioning of thin-film pv modules through controlled light...
Post on 22-Aug-2020
3 Views
Preview:
TRANSCRIPT
Preconditioning of Thin-Film PV Modules Through Controlled Light-Soaking
Tony SampleDG-JRC Ispra, Italy
�• Based in Ispra, northern Italy
Overview
�• General introduction to meta-stabilities in Thin-Film devices�• Some examples of what has been observed in the
literature
�• Methods used to stabilize Thin-Film devices�• Light-soaking according to IEC 61646 ed 2 (2008)
�• Outline of light-soaking experiments with various Thin-Film modules�• Modules used in the experiments�• Results �• Conclusions from the experiments
�• Overall Conclusions
Amorphous Silicon, including Tandem, micromorph and triple junction
�• Exhibit a long-term meta-stable behaviour in which their maximum power decreases with light exposure but improves through thermal annealing, the Staebler-Wronskieffect [1]
�• Micromorph silicon (a-Si/ -Si) materials are also affected because they contain an amorphous layer, but it is more stable than single junction amorphous silicon [2,3]
�• All exhibit a slow decrease in power due to cold soaking and a much faster recovery through thermal annealing [2,3]
�• The impact of the different time constants is reflected in the observed seasonal variations [2-4]
�•
[1] D. L. Staebler, C. R. Wronski, 1997[2] J. A. del Cueto, and B. von Roedern 1999.
[3] M. Nikolaeva-Dimitrova et al 2010 [4] D.L. King et al 2000
Copper Indium Gallium (di)Selenide (CIGS)
�• CIGS modules are also subject to light-induced change of the module efficiency [5].
�• Not as clear behaviour pattern for CIGS�• Some authors have shown that they may degrade [6, 7] with light
exposure �• but in some cases it has been shown that they remain stable [7] or
improve [8].
�• The behaviour is very dependent on the deposition and exact material composition.
�• In general these modules exhibit a short-term meta-stable behaviour modulated by light and for this reason they have to be measured quickly following light exposure [8]
[5] A. G. Aberle, 2009[6]. E. L. Meyer and E. E. van Dyk, 2003
[7] C. Radue et al 2009 [8] R. P. Kenny et al 2006
Cadmium Telluride (CdTe)
�• Early generation CdTe modules exhibited a long-term meta-stable behaviour but modules could either degrade or improve with light exposure [9]
�• However, recent modules have been shown a more uniform behaviour in that they tend to increase their maximum power with light exposure, especially following storage in the dark [9-11]
[9] J. A. del Cueto and B. von Roedern 2006[10] First Solar, Inc. 2009.
[11] Z. Jingquan et al 2009
Methods used to stabilize Thin-Film devices
�• Some groups have experimented with the use of current injection to stabilize CIGS devices, in particular to overcome the very fast degradation on dark storage. However, this approach will not be detailed here.
�• Light-soaking according to IEC 61646 ed 2 (2008) [12]Stabilization occurs when measurements from two consecutive periods of at least 43kWh/m2 each integrated over periods when the temperature is between 40°C and 60°C, meet the following criteria:
(Pmax – Pmin) / Paverage < 2%
[12] IEC 61646 2008
Light-soaking apparatus
�• Large climatically controlled chamber containing a class BBB solar simulator (on the limit for spectral match CBB)
Irradiance: 850-870 W/m2
Duration per period ~48 Hours
Module temperature 45-55oCStability better than ±2oC
Operation under resistive load
Module Pmax determined on class AAA simulator at 25oC
Power measurements
�• The IV characteristics are measured by sweeping the device from Isc to Voc using a SpectroLab X25 LAPSS has a light pulse of 2 ms duration with a flat irradiance of 1000 W/m2.�• Measured IV characteristics according to IEC 60904-1. It is noted that
this standard may be applicable to multi-junction test specimens, if each sub-junction generates the same amount of current as it would under the reference AM1,5 spectrum in IEC 60904-3.
�• It is assumed that the spectral mismatch between the various modules and the reference cell remains constant throughout the repeated light exposures.
�• It is also assumed that the limiting junction of the multi-junction devices does not switch due to light-soaking when measured using the solar simulator
�• Relative uncertainties of �• Isc ±1.3%, Voc ±0.3%, FF ±0.72%, Pmax ±1.5% [13]
[13] H. Muellejans et al 2009
Crystalline Silicon on Glass (CSG)NW74
a-Si/ -SiHJ410
a-Si/ -SiKW712
a-Si/ -SiKW711
Triple junction a-SiLK712
Triple junction a-SiLK711
CdTeNW73
CdTeKX711
CIGS (Copper-Indium-Gallium-diselenide)NW71
CIGS (Copper-Indium-Gallium-diselenide)LF711
CIGS (Copper-Indium-Gallium-diselenide)BY71
TECHNOLOGYESTI CODE
Modules used in the study
Note: Modules with the same two letter code are from the same manufacturer and batch.The modules had taken part in different projects in the past and have therefore had varying histories of light and temperature exposure. After the end of these previous projects they were stored in the dark, near to 25°C. The length of storage is different for each one of them and varies from several weeks to several months.
Results for CIGS modules
Modules measured within 30 minutes of leaving the light-soaking chamber
All modules exhibit an increase of Pmax with light-soaking
Stabilised after 3-4 exposure cycles
Dark storage leads to a decrease in Pmax
Note:
Open symbols indicate a faulty connection leading to a lower than expected power
Results for CdTe modules
Light-soaking has improved the maximum power
Difference between the two modules due to their prior history
Continued to improve following point of stabilisation
Results for a-Si/µ-Si and Triple junction
Light-soaking has improved the maximum power
The two triple junction modules exhibited a smaller decrease in power than the a-Si/µ-Si modules
Continued to decrease following point of stabilisation
Results for CSG
CSG material clearly not meta-stabile
No need to be subjected to light-soaking
For the purposes of module qualification, the stability procedure of IEC 61646 ed 2 is probably satisfactory, given the need to stay “within reasonable constraints of cost and time”.
For a-Si containing modules this will tend to lead to an over estimation of the power outputFor CdTe modules this would tend to lead to an underestimation Valid for CIGS
not required for CSG
For thin-film module calibration, in general applying the stability procedure of IEC 61646 is not sufficient, therefore:
More stringent stability criteria are suggested, such as more periods of light soaking and/or tighter stability limits.The a-Si community has tended to use 1000 hours @ 1 Sun
Conclusions
One aspect of the stabilization process not explicitly studied here, but worthy of further examination is the choice of irradiance level and temperature. The standard calls for;
A class CCC solar simulator, in accordance with the IEC 60904-9, or natural sunlight..consecutive periods of at least 43 kWh m�–2 each integrated over periods when the temperature is between 40°C and 60°C, meet the following criteria: (Pmax �– Pmin)/Paverage < 2 %.
�• If using controlled indoor light-soaking you could choose a target temperature from 40 to 60oC
Amorphous silicon devices will have the greatest maximum output at the highest temperature
For outdoor light-soaking under natural sunlightNo control of temperatureThe integrated exposure can be calculated, but what happens if the module exceeds 60oC ?
Conclusions
References[1] D. L. Staebler, C. R. Wronski, �“Reversible conductivity changes in discharge-produced amorphous
Si�”, Appl. Phys. Lett. 31, pp 292-294 [2] J. A. del Cueto, and B. von Roedern �“Temperature-induced Changes in the Performance of
Amorphous Silicon Multi-junction Modules in Controlled Light- soaking�” Prog. Photovolt: Res. Appl. 7, 101-112 (1999)
[3] M. Nikolaeva-Dimitrova, R. P. Kenny, E. D. Dunlop, "Controlled Conditioning of a-Si:H Thin Film Modules for efficiency prediction", Thin Solid Films, Vol. 516/20, 2008, pp. 6902-6906.
[4] D.L. King, J. A. Kratochvil, and W. E. Boyson �“Stabilization and Performance Characteristics of Commercial Amorphous-Silicon PV Modules�” Proceeding of the 28th IEEE PVSC Anchorage USA, September 2000, pp. 1446-1449
[5] A. G. Aberle, �“Thin-film solar cells�”, Thin Solid Films, Vol. 517, No. 17, July 2009, pp. 4706-4710.[6] E. L. Meyer and E. E. van Dyk, �“Characterization of degradation in thin-film photovoltaic module
performance parameters�”, Renewable Energy, Vol. 28, No. 9, July 2003, pp. 1455-1469.[7] C. Radue, E. E. van Dyk, E. Q. Macabebe, �“Analysis of performance and device parameters of CIGS
PV modules deployed outside�”, Thin Solid Films, Vol. 517, No. 7, February 2009, pp. 2383-2385.[8] R. P. Kenny, M. Nikolaeva-Dimitrova, E. D. Dunlop, �“Performance Measurements of CIS Modules:
Outdoor and pulsed simulator Comparison for Power and Energy Rating�”, Proceedings of the 4th WCPEC, Hawaii, May 2006. pp. 2058-2061.
[9] J. A. del Cueto and B. von Roedern �“Long-term Transient and Metastable Effects in Cadmium Telluride Photovoltaic Modules�” Prog. Photovolt: Res. Appl. 2006; 14:615�–628
[10] First Solar, Inc., �“Application Note: Best Practise for Power Characterization�”, PD-5-434, Rev 1.0, downloaded from www.firstsolar.com, 2009.
[11] Z. Jingquan F. Lianghuan, L. Zhi, C. Yaping, L. Wei, W. Lili, L. Bing, C. Wei, Z. Jiagui, �“Preparation and performance of thin film CdTe mini-module�”, Solar Energy Materials and Solar Cells, Vol. 93, No. 6-7, June 2009, pp. 966-969.
[12] IEC 61646. Thin-film terrestrial photovoltaic (PV) modules - Design, qualification and type approval (2nd ed.). IEC Central Office: 2008.
[13] H. Muellejans, W. Zaaiman, R. Galleano, �“Analysis and mitigation of measurement uncertainties in the traceability chain for the calibration of photovoltaic devices�”, Meas. Sci. Technol., 20, (2009): 075101 (12pp).
�• The work was partially funded by the PERFORMANCE project of the European Commission under contract number SES-019718 within FP6.
Anatoli Chatzipanagi, Robert Kenny, Mike Field and Ewan Dunlop
Acknowledgments
top related